There has been strong enthusiasm to adopt artificial intelligence (AI) within the healthcare sector, from diagnostics to novel drug and clinical trial selection.
The Australian College of Nursing (ACN) has thrown its hat into the AI arena, releasing a Position Statement outlining the ACN’s stance on the safe, ethical use of AI in nursing practice, education and administration.
ACN interim CEO Emeritus Professor Leanne Boyd FACN said nurses were uniquely placed to lead in developing, testing, implementing and evaluating AI in healthcare.
“While AI has many potential benefits in healthcare, appropriate regulations and safeguards must be embedded to not compromise patient safety, nursing care delivery or the profession more broadly,” Professor Boyd said.
“AI has the potential to significantly reduce the often-repetitive tasks that nurses perform, as well as assist in solving both our current and future workforce challenges.”
ACN recommendations for AI
The ACN’s recommendations are designed to serve as a guide for nurses while acknowledging the role AI will play in the contemporary healthcare setting.
The college’s priority is to maintain a patient and nurse-centred decision-making structure in healthcare, ensuring nurses remain the final decision maker, leveraging their knowledge, training and critical thinking in care.
To that end, the ACN advocates for AI training at all levels of nursing to ensure a comprehensive understanding of available products, algorithmic decision-making and the legal liabilities associated with automated decisions.
“Nurses must play a central role in designing, implementing and evaluating AI applications, ensuring that ethical and practical considerations align with nursing requirements,” ACN's statement reads.
“AI should only be integrated into nursing practice when ratified evidence demonstrates improved patient outcomes.
“It is imperative to emphasise that AI is a tool to enhance nursing care and treatment, not a replacement for critical thinking.”
To ensure that outcome, the ACN believes nurses should be actively involved in developing data governance models based on principles of integrity, transparency, auditability, accountability, stewardship, checks and balances, standardisation and change management.
“We advocate for the recommendations outlined in A National Policy Roadmap for Artificial Intelligence in Healthcare (AAAiH, 2023),” the ACN statement says.
“We advocate for AI to be developed within a robust safety framework to implement accreditation to assess AI safety and quality practice standards and integrate the national AI ethical framework to support value-based healthcare.”
AI already transforming healthcare
The use case for AI in diagnostics is already gaining momentum as companies leverage the technology in early detection, particularly of cancer.
Breast cancer screening is a massive industry, valued at US$7.96 billion in 2023. Data from the American Cancer Society indicates breast screens in the US can generate high levels of false positives, up to 11%.
The use of AI in imaging analysis has increased accuracy materially, allowing mammograms to be interpreted 30-times faster with 99% accuracy.
While false positives for breast cancer screening are much lower in Australia, about 3.8%, the benefits of rapid analysis and even higher levels of accuracy are obvious.
A report by MarketsandMarkets estimates the global AI in healthcare market will reach US$67.4 billion by 2027, expanding at a compounding annual growth rate (CAGR) of 46%.
The Australian Computer Society reckons a specialist workforce of between 32,000 and 161,000 individuals skilled in AI will be necessary by 2030.
AI is already being used to encourage better lifestyle choices, increase accuracy of disease diagnosis, facilitate data-based decision-making, improve treatment plans and to enable new types of research and education.
AI has the potential to transform the accuracy, speed and efficacy of the healthcare system but only if it’s tempered by human critical thinking and empathy – the technology is altogether too prone to bias without oversight.